Film based techniques have generally been investigated for a wide variety of sensors, as reported by Wenyi et al., 1997; Hughes et al., 1997; Staley, 1996; Agbor et al., 1995; Tan and Tan, 1995; Menil et al., 1994; Kunnecke et al., 1994; Creasey and Varney, 1994; Geistlinger, 1993; Ishiji et al., 1993; Najafi et al., 1992; Hampp et al., 1992; Nakano and Ogawa, 1994; Yamazoe and Miura, 1994; and, Madou and Otagawa, 1989. While solid-state gas sensors may have the advantage of being able to operate at elevated temperatures, they also may have the disadvantages of slow response and recovery time and a high internal operating temperature as reported by Liu et al., 1993; and Narducci et al., 1993. More recent literature (Schwebel et al., 1997; Sheng et al., 1997; Micocci et al., 1997) eludes to more substantial development work yet to be done.
In Kinlen et al., 1994, a Nafion®-coated metal oxide pH sensor is generally characterized as having sputtered iridium oxide sensing and silver/silver chloride reference electrodes on alumina ceramic substrates. Nafion may have been used as a cation-selective ionomer coating in order to decrease the oxidation-reduction error typically affecting the performance of metal oxide pH electrodes. In Yasuda et al., 1994, the use of Nafion as polymer-electrolyte for a thin-film CO sensor is generally described with macro-sized, sputtered Pt sensing and counter electrodes and a smaller, sputtered Au electrode as a reference electrode. A 5 wt % n-propyl alcohol solution of Nafion (DuPont, 1100 EW) may be used to form the polymer electrolyte film over the electrodes by casting. The polymer is usually washed and protonated in aqueous sulfuric acid prior to casting. A theorized lifetime of this sensor is normally less than one month. During this time, the CO oxidation current typically decreases steadily down to a few percent of its original value without any period of stable measurement signal. The lifetime of the device may be extended up to three years by laminating the polymer electrolyte layer with a cast perfluorocycloether-polymer film in order to keep the CO permeability coefficient through Nafion constant. Theoretical calculations often reflect the drift rate of the signal could be significantly reduced under these conditions.
A description of typical state-of-the-art hydrated solid polymer electrolyte or ionomer sensors and sensor cells is generally described by Kosek et al. U.S. Pat. No. 5,527,446; LaConti and Griffith, U.S. Pat. No. 4,820,386; Shen et al., U.S. Pat. No. 5,573,648; and, Stetter and Pan, U.S. Pat. No. 5,331,310. These sensor cells, based on hydrated solid polymer electrolyte or ionomer technology, may have several advantages over conventional electrochemical sensor cells. The catalytic electrodes are normally bonded directly to both sides of a proton conducting solid polymer ionomer membrane providing a stable electrode to electrolyte interface. One side of the electrolyte membrane is usually flooded with distilled water, making the sensor cell self-humidifying and independent of external humidity. Since no corrosive acids or bases are generally used in the sensor cell, a lifetime of over 10 years may be experienced for solid polymer ionomer sensor cells. Finally, the sensor cells may be easy to maintain and may be ideal for use in remote, unattended environments because maintenance typically entails little more than addition of water to the reservoir in the sensor housing every several months and monthly calibration checks.
A disadvantage of the state-of-the-art sensors described above may be that the signal-to-noise ratio is not be conducive to detection of very low concentrations (parts per billion, ppb) of important environmental and biomedical gases and vapors. Also, response time may be relatively slow, and reproducibility between sensors and sensor cells may be difficult to achieve. Also, they are relatively costly.
Recently, miniaturized thick- and thin-film type sensors have been developed where the solid ionomer membrane often acts as a conduit between the gas to be detected (sample gas) and the sensing electrode (Yasuda et al., 1994). The sample gas usually permeates through the membrane itself where a 3-phase contact area is established. A disadvantage with this configuration may be that the solid ionomer membrane water content often controls the gas permeation rate as well as proton conductivity. As the humidity increases, the membrane water content typically increases. This may cause an increase in the gas diffusion rate as well as proton conductivity and sensor signal response. A method for controlling or fixing the water content of the membrane may be to have a water reservoir on the back side of the membrane, directly opposite to where the film type electrodes and non-conductive supportive substrate are located. However, the back side of the membrane is often required to be free of liquid so that the sample gas can diffuse through the membrane to the sensing electrode.
U.S. Pat. No. 4,812,221 to Madou et al. (“Madou”) typically relates to a gas sensor having a porous member located in a passage adjacent and generally in contact with a sensing electrode. The pore size of the porous member may be controlled by varying the processing parameters, such as current, hydrogen fluoride concentration, and the like. In addition, Madou appears to indicate a number of other steps for providing the gas sensor, such as sizing the pores in the porous membrane, sizing the pores of the sensing electrode, and selecting the materials of the electrodes. A problem often associated with a sensor provided in accordance with Madou is difficulty in repeatability and/or reproducibility due to the numerous variations from one sensor to another. Another difficulty may be problems or costs in manufacturing sensors due to the quantity of steps, where it is often believed that manufacturing becomes more expensive as the quantity of steps is increased. Another possible disadvantage is that the permeability coefficient of the filter in Madou is not disclosed to be utilized for determining the membrane's thickness and other physical characteriztics, which optimizes the sensor's response time. Still a further possible disadvantage is that Madou's filter is inert and does not react, if desired, with the gas that diffuses through the membrane.
What is desired, therefore, is a sensor having improved repeatability. Another desire is to provide a sensor having improved response time. A further desire is to provide a sensor having an improved signal to noise ratio. A still further desire is to provide a sensor that is easy to manufacture with reduced costs while maintaining or improving its operational and manufacturing efficiency.